Claim · model-written summary
The classification accuracy of CSNet was 13.3% higher than that of RF.
Not reportedTaxonMaterialGeneCompoundOrgan / stageInteractorInteractor roleSecond interactorSecond interactor roleTreatmentSettingQualifier
Verbatim
e4“The classification accuracy is 13.3% and 9.5% points higher than RF and SVM, respectively”
Title
Fine crop classification in high resolution remote sensing based on deep learning
Tingyu Lu; Luhe Wan; Lei Wang Frontiers in Environmental Science 2022. doi:10.3389/fenvs.2022.991173
Other claims from this paper (7)
- Method / resource The highest classification accuracy of CSNet based on fine-tuned Resnet-50 was 91.2%.
- Method / resource The kappa coefficient of CSNet was 0.882.
- Method / resource The mean intersection over union of CSNet was 0.834.
- Method / resource The classification accuracy of CSNet was 9.5% higher than that of SVM.
- Method / resource Fusing multi-spectral information with vegetation index features helps improve classification accuracy.
- Method / resource The deep learning algorithm is superior to the machine learning algorithm in classification accuracy and classification effect.
- Method / resource The CSNet model exhibits good transferability.